EPA 8260, chromatographic problem after possible foamer

[James_Ball]

Ok, so we one of the analyst had a sample last week that we think foamed on the purge and trap. A filament was blown and I had to retune since the 502 mass was about 2amu wide, which sounds like a little junk got into the analyzer.

Tune looks good, first couple chromatograms looked like total junk. So I changed the trap and it got better but still terrible. Cleaned ever line in the system that I could not replace, and replaced not only the transfer line but also some in the valve oven with silcosteel line. Also changed column and cleaned injection port. Still I have a problem where Pentafluorobenzene, 1,4-Difluorobenzene and Toluene-d8 are giving broad peaks. At least now they look gaussian, where before Pentafluorobenzene looked like a ski slope, then became a broad hump then a peak that looks like it is being run on a packed column at best. Weird thing is, 1,2-Dichloroethane falls between Pentafluorobenzene and 1,4-Difluorobenzene and it gives a nice sharp peak.

I eliminated any connections that could have given dead volume also, just to make certain that was not the problem. It had perfect peak shapes before the sample, so it is related to the bad sample. I haven't opened it up and cleaned the source yet, but I would not think that would effect peak shapes this much. I am running a series of blanks tonight to see if it cleans up any more so wish me luck.

If anyone has seen something like this before it would be great to hear your resolution to the problem. There has to still be some contamination somewhere but I just can't seem to find it.

The instrument is a 5973/6890 with Encon purge and trap running a Vocarb3000 trap(second one was taken from another instrument that was working great). Column is a Restek Rtx502.2 40m x 0.18 1um film and is also new.

[Yama001]

I would wonder if the pneumatics module on the 6890 has been fouled. It might make sense to swap it with one from another 6890 if you have one. Those modules have a split line filter, but a bad foamer can reach it through the septum purge pathway, or even overwhelm the split filter ( I assume you have changed that). Extensive damage here would be likely prevent the inlet from even reaching a ready state, but perhaps you are just lucky enough to have an apparently functional but damaged module .

I suppose it is possible water and gunk left in the flow path somewhere has damaged the new column and/or trap. I usually verify columns by doing a direct injection and looking at the ketone peak shapes.

It might be worth swapping another concentrator in to see if the issue is the GC/MS verse something that is still damaged in the original concentrator. I get worried about the alternate flow paths that are used for dry purge, bake gas bypass, backfush desorb and bake, etc. I am not sure if the Encons have recommended values for the trap back pressure at various stages in the cycle, but that may be useful info.

Rick Haney at EST is pretty sharp on this stuff, probably worth a call.

[James_Ball]

I have worked with Rick, Big Dan, and several others at EST, if I can't get it today that is my next call. Maybe just time to get rid of this old one and upgrade to and Evolution

I just swapped in another trap, clipped some column and replaced the split vent line. Funny I noticed that the injection port end of the split line was silver from the ferrule into the inlet, the rest of the line was regular copper color, and it was a little charred looking on the inside, but I normally swap those out about once a year as it is. The silver on the tip is worrying me though, I really really hope it is not mercury plated out on the copper, that would be about the end of the instrument. Had that happen once before about 15 years ago, it took out the analyzer on a 5971 when someone sent me a sample of carbon contaminated with mercury and we ran the volatiles first, then found out about the mercury the next day after metals was run on it.

Will know soon if I finally got it cleaned out. Crossing my fingers )

[Bigbear]

I had a similar situation where someone sent me incenerator distillate marked as drinking water! It even trashed my column.
Have you tried a direct injection of your internals and surrogates?

[James_Ball]

Direct injections looked like perfect peaks. If I injected onto the trap then ran a blank purge it would give a sharp peak for the internals followed by a blob peak of the same internal, but only the ones that were problems. I replaced almost every piece of line I could and even rerouted some lines to bypass parts of the purge and trap with no luck. I even moved the transfer line so it went directly through the septa instead of the incoming gas line on the injection port just in case.

Dan from PTS was here today and did a refurb, replacing all the internal plumbing on it, and now I have perfect peaks and great response again.

Second time these sample have killed an instrument in a year. I hate when clients ask for ultra trace results on samples that are loaded with junk. Ran another last night on another instrument that I had to dilute 200,000x just to keep the targets below 400ppb on the instrument lol.

[Bigbear]

Great that you are back up and running. As I remember when it happened to us, we charged the customer for most of the repair costs.

Do you have a static headspace setup you could screen suspect samples on?

[James_Ball]

Great that you are back up and running. As I remember when it happened to us, we charged the customer for most of the repair costs.

Do you have a static headspace setup you could screen suspect samples on?

We have a static headspace on an FID but it is usually busy running other samples.

High analyte concentration didn't kill this one I am afraid, I believe it foamed and carried non-volatile residues over into the plumbing. Usually a steam cleaning or flush with methanol will fix it, but this time the crud just would not come out.

I would charge the client, but we know what the samples do, and this was more of a new analyst not realizing what the sample was, type of situation. So it lies on us on this one.

The EPA has told the client they must run it by 8260 for their list, but we did win one victory there when we convinced the EPA that you could not get the low detection limits they were wanting(18ppb) for Acrylic Acid by purge and trap 8260. This stuff looks like Elmer's Glue and is sometimes thicker than glue and they were wanting normal water detection limits on it

I wish the EPA had to run it on their equipment and see what happens lol.

[MichaelVW]

Funny I noticed that the injection port end of the split line was silver from the ferrule into the inlet, the rest of the line was regular copper color, and it was a little charred looking on the inside, but I normally swap those out about once a year as it is. The silver on the tip is worrying me though, I really really hope it is not mercury plated out on the copper, that would be about the end of the instrument.

Is that common? (I have nothing to add here, I'm just curious!) Does it have to be a killer sample or might it build up over time from the few ppm range?
Any other metals in particular that need looking out for?

[James_Ball]

Funny I noticed that the injection port end of the split line was silver from the ferrule into the inlet, the rest of the line was regular copper color, and it was a little charred looking on the inside, but I normally swap those out about once a year as it is. The silver on the tip is worrying me though, I really really hope it is not mercury plated out on the copper, that would be about the end of the instrument.

Is that common? (I have nothing to add here, I'm just curious!) Does it have to be a killer sample or might it build up over time from the few ppm range?
Any other metals in particular that need looking out for?

Not sure how bad it would be from a slow buildup over time. I would hope that most would stay in the gas phase and get pumped out, but that is probably wishful thinking. The one that killed ours was probably percent level mercury trapped in carbon, not as bad as the one we ran in metals once that took days to flush out and when we poured the water out of the bucket the sample came in we found liquid mercury on the bottom. Needless to say, that one got sent back to the client for disposal.

[Yama001]

I usually end up replacing the (somewhat) affordable parts in the flow path when it gets this bad - basically all the plumbing except the 6 port valve and the sample mount.

I am wondering what else was covered in the EST refurb? Did they replace the transfer line to the GC, or any of the valves/tees that serve for dry purge, backflush, bypass, etc?

Thanks.

[James_Ball]

I usually end up replacing the (somewhat) affordable parts in the flow path when it gets this bad - basically all the plumbing except the 6 port valve and the sample mount.

I am wondering what else was covered in the EST refurb? Did they replace the transfer line to the GC, or any of the valves/tees that serve for dry purge, backflush, bypass, etc?

Thanks.

Refurb included the injection ports on the front(it is the older Encon model) all the tubing inside, I believe also the mount, switching valve, trap heater, MORT heater, all the lines going to those on top and bottom and the transfer line to the GC. I think it was everything that a sample would pass through except the soil side transfer line from the Archon to the Encon(that is covered in an Archon refurb).

It is less expensive to ship it to them and have it done there, but we had them come out and do it here to save shipping time back and fourth. It was finished and up and running before 2pm, so only a few hours to complete.

[Yama001]

Thanks James.